Name: Automated annotation of epileptiform burden and its association with outcomes
Published: July 11, 2026
License: https://github.com/bdsp-core/bdsp-license-and-dua

Software Credentialed Access

Sahar F. Zafar , Eric Rosenthal , Jin Jing , Wendong Ge , Mohammad Tabaeizadeh , Hassan Aboul Nour , Maryum Shoukat , Haoqi Sun , Farrukh Javed , Solomon Kassa , Muhammad Edhi , Elahe Bordbar , John Gallagher , Valdery Moura , Manohar Ghanta , Yu-Ping Shao , Sungtae An , Jimeng Sun , Andrew J. Cole , M. Brandon Westover

Published: July 11, 2026. Version: 1.0.0

When using this resource, please cite: (show more options)
Zafar, S. F., Rosenthal, E., Jing, J., Ge, W., Tabaeizadeh, M., Aboul Nour, H., Shoukat, M., Sun, H., Javed, F., Kassa, S., Edhi, M., Bordbar, E., Gallagher, J., Moura, V., Ghanta, M., Shao, Y., An, S., Sun, J., Cole, A. J., & Westover, M. B. (2026). Automated annotation of epileptiform burden and its association with outcomes (version 1.0.0). Brain Data Science Platform. https://doi.org/10.60508/qv17-ps72.

MLA	Zafar, Sahar F., et al. "Automated annotation of epileptiform burden and its association with outcomes" (version 1.0.0). Brain Data Science Platform (2026), https://doi.org/10.60508/qv17-ps72.
APA	Zafar, S. F., Rosenthal, E., Jing, J., Ge, W., Tabaeizadeh, M., Aboul Nour, H., Shoukat, M., Sun, H., Javed, F., Kassa, S., Edhi, M., Bordbar, E., Gallagher, J., Moura, V., Ghanta, M., Shao, Y., An, S., Sun, J., Cole, A. J., & Westover, M. B. (2026). Automated annotation of epileptiform burden and its association with outcomes (version 1.0.0). Brain Data Science Platform. https://doi.org/10.60508/qv17-ps72.
Chicago	Zafar, Sahar F., Rosenthal, Eric, Jing, Jin, Ge, Wendong, Tabaeizadeh, Mohammad, Aboul Nour, Hassan, Shoukat, Maryum, Sun, Haoqi, Javed, Farrukh, Kassa, Solomon, Edhi, Muhammad, Bordbar, Elahe, Gallagher, John, Moura, Valdery, Ghanta, Manohar, Shao, Yu-Ping, An, Sungtae, Sun, Jimeng, Cole, Andrew J., and M. Brandon Westover. "Automated annotation of epileptiform burden and its association with outcomes" (version 1.0.0). Brain Data Science Platform (2026). https://doi.org/10.60508/qv17-ps72.
Harvard	Zafar, S. F., Rosenthal, E., Jing, J., Ge, W., Tabaeizadeh, M., Aboul Nour, H., Shoukat, M., Sun, H., Javed, F., Kassa, S., Edhi, M., Bordbar, E., Gallagher, J., Moura, V., Ghanta, M., Shao, Y., An, S., Sun, J., Cole, A. J., and Westover, M. B. (2026) 'Automated annotation of epileptiform burden and its association with outcomes' (version 1.0.0), Brain Data Science Platform. Available at: https://doi.org/10.60508/qv17-ps72.
Vancouver	Zafar S F, Rosenthal E, Jing J, Ge W, Tabaeizadeh M, Aboul Nour H, Shoukat M, Sun H, Javed F, Kassa S, Edhi M, Bordbar E, Gallagher J, Moura V, Ghanta M, Shao Y, An S, Sun J, Cole A J, Westover M B. Automated annotation of epileptiform burden and its association with outcomes (version 1.0.0). Brain Data Science Platform. 2026. Available from: https://doi.org/10.60508/qv17-ps72.

Additionally, please cite the original publication:

Zafar SF, Rosenthal ES, Jing J, Ge W, Tabaeizadeh M, Aboul Nour H, Shoukat M, Sun H, Javed F, Kassa S, Edhi M, Bordbar E, Gallagher J, Moura VJ, Ghanta M, Shao YP, An S, Sun J, Cole AJ, Westover MB. Automated Annotation of Epileptiform Burden and Its Association with Outcomes. Ann Neurol. 2021;90(2):300-311. PMID: 34231244.

Abstract

Objective. Ictal-interictal continuum (IIC) and seizure activity are common on continuous EEG in critically ill patients, but the relationship between the burden of this activity and outcome has been hard to quantify at scale. Using automated annotation, we measured epileptiform burden in 1,991 patients and related it to discharge outcome.

Methods. Epileptiform/IIC and seizure burden were quantified from continuous EEG by an automated pipeline. Burden and clinical covariates were related to modified Rankin Scale (mRS) outcome using nested cross-validated LASSO logistic regression, with ROC/calibration, covariate-adjusted dose-response, and subgroup analyses.

Results. Higher epileptiform burden was independently associated with worse outcome; the covariate+burden model predicts poor outcome with good discrimination (cross-validated AUC ~0.80). This project releases the de-identified 1,991-patient analysis data and code.

Background

Continuous EEG frequently shows ictal-interictal-continuum (IIC) patterns and seizures in critically ill patients. Quantifying the burden of this activity and its association with outcome, at scale, requires automated annotation. This project releases the data and code from a 1,991-patient study.

Software Description

MATLAB analysis organized by figure/table (feature selection, ROC/calibration, dose-response, burden swimmer, predictor statistics) and de-identified data: InputArray_pred_outcome.mat (1,991 x 16 predictors + mRS outcome), selected-feature and per-figure .mat files. Numeric only; no names, MRNs, or dates.

Technical Implementation

Epileptiform/IIC and seizure burden were automatically quantified from continuous EEG and related to modified Rankin Scale outcome via nested cross-validated LASSO logistic regression, with ROC/calibration, covariate-adjusted dose-response, and subgroup analyses.

Installation and Requirements

MATLAB. Run each figure/table folder's main_* script (scripts use Windows path separators; on macOS/Linux replace backslashes with forward slashes). See REPRODUCE.md and DATA_SOURCE.md.

Usage Notes

The committed InputArray_pred_outcome.mat reproduces the burden->outcome prediction (10-fold CV AUC ~0.80, poor outcome = mRS>=4). main_featureSelection_LASSO_nestedCV.m fits the model; the Figure*/Table2 scripts regenerate the paper's items.

Release Notes

First public release: de-identified 1,991-patient burden/outcome data + analysis code.

Ethics

De-identified data under IRB approval; numeric analysis matrices only, no identifiers.

Acknowledgements

Data from the MGH continuous-EEG cohort.

Conflicts of Interest

See the associated publication (Ann Neurol 2021;90:300-311).

References

Zafar SF, Rosenthal ES, Jing J, Ge W, Tabaeizadeh M, Aboul Nour H, Shoukat M, Sun H, Javed F, Kassa S, Edhi M, Bordbar E, Gallagher J, Moura VJ, Ghanta M, Shao YP, An S, Sun J, Cole AJ, Westover MB. Automated Annotation of Epileptiform Burden and Its Association with Outcomes. Ann Neurol. 2021;90(2):300-311. PMID: 34231244.